using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

8
This article was downloaded by: [Universitat Politècnica de València] On: 25 October 2014, At: 11:13 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK International Journal of Remote Sensing Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tres20 Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps Einar Bjorgo Published online: 25 Nov 2010. To cite this article: Einar Bjorgo (2000) Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps, International Journal of Remote Sensing, 21:3, 611-616, DOI: 10.1080/014311600210786 To link to this article: http://dx.doi.org/10.1080/014311600210786 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

Upload: einar

Post on 28-Feb-2017

217 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

This article was downloaded by: [Universitat Politècnica de València]On: 25 October 2014, At: 11:13Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number:1072954 Registered office: Mortimer House, 37-41 Mortimer Street,London W1T 3JH, UK

International Journal ofRemote SensingPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/tres20

Using very high spatialresolution multispectralsatellite sensor imagery tomonitor refugee campsEinar BjorgoPublished online: 25 Nov 2010.

To cite this article: Einar Bjorgo (2000) Using very high spatial resolutionmultispectral satellite sensor imagery to monitor refugee camps, InternationalJournal of Remote Sensing, 21:3, 611-616, DOI: 10.1080/014311600210786

To link to this article: http://dx.doi.org/10.1080/014311600210786

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of allthe information (the “Content”) contained in the publications on ourplatform. However, Taylor & Francis, our agents, and our licensorsmake no representations or warranties whatsoever as to the accuracy,completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views ofthe authors, and are not the views of or endorsed by Taylor & Francis.The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor andFrancis shall not be liable for any losses, actions, claims, proceedings,demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, inrelation to or arising out of the use of the Content.

Page 2: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

This article may be used for research, teaching, and private studypurposes. Any substantial or systematic reproduction, redistribution,reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of accessand use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 3: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

int. j. remote sensing, 2000, vol. 21, no. 3, 611–616

Using very high spatial resolution multispectral satellite sensorimagery to monitor refugee camps

EINAR BJORGO

Nansen Environmental and Remote Sensing Centre, Edvard Griegsvei 3A,N-5059 Bergen, Norway

(Received 20 October 1998; in � nal form 19 May 1999)

Abstract. Detailed geographic information is a key factor in decision makingprocesses during refugee relief operations. The forthcoming commercial very highspatial resolution (VHSR) satellite sensors will be capable of acquiring multi-spectral (MS) images at spatial resolutions of 1 m (panchromatic) and 4 m(multispectral ) of refugee camps and their environment. This work demonstrateshow refugee camp environment, area and population can be analysed using aVHSR MS satellite sensor image from the Russian KVR-1000 sensor. This image,with a spatial resolution of 3.3 m, was used to study Thailand’s Site 2 refugeecamps, which were established to accommodate Khmer refugees on theThai–Kampuchean border. At the time of image acquisition, the total populationof Site 2’s � ve refugee camps was close to 143 000. The VHSR MS image wasfound to be suitable for mapping the refugee camp environment and area. Astatistically signi� cant linear relationship between camp area and population wasdetermined. Accordingly, the study suggests that VHSR MS images in generalmay be useful for refugee camp planning and management and points toward theutilization of forthcoming commercial VHSR MS satellite sensor images inhumanitarian relief operations.

1. IntroductionHumanitarian organisations need reliable and up-to-date information about the

situation on the ground for e� cient planning and management of relief operations(Bouchardy 1995, Elo 1996). Satellite images can provide objective and updatedinformation about the environment in which refugee camps are placed (Bouchardy1995, Gupta 1995). Although satellite images have been used for regional-scalemapping (Lodhi et al. 1998), most relief agencies have not used this technologyextensively. It is expected that images from the forthcoming very high spatial reso-lution (VHSR) satellite sensors, onboard polar-orbiting platforms such as Ikonos-2,QuickBird, Orbview-3, IRS-P6, and EROS A&B (as described by Aplin et al. 1997 ),will have less need for specialist interpretation and will thus be easier for non-specialist relief sta� to analyse. VHSR satellite sensors are commonly de� ned by theremote sensing community to provide imagery below the 4 m spatial resolution limit.Single-channel, panchromatic VHSR images have proved useful for mapping andfeature recognition (Gupta 1995). For vegetation assessment and change detectionmultispectral (MS) colour images are generally more suitable than panchromaticimages (Tucker 1979).

Internationa l Journal of Remote SensingISSN 0143-1161 print/ISSN 1366-5901 online © 2000 Taylor & Francis Ltd

http://www.tandf.co.uk/journals/tf/01431161.html

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 4: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

E. Bjorgo612

This Letter demonstrates how VHSR MS satellite sensor imagery can be usedto derive detailed information on refugee camp environment, area and population.It is based on analysing recently released archival VHSR MS satellite sensor imageryacquired from Russian satellites.

2. Study area and dataThe now demolished refugee camp surveyed in this case study, Thailand’s Site 2,

was located approximately 80 km north of the town of Aranyaprathet in thePrachinburi province bordering Cambodia and peaked at a population of 143 000refugees (Lynch 1990). Site 2 was selected for this case study because both in situ

refugee camp information (Lynch 1990, Shook 1990, R. Burrows 1998, Personalcommunication) , and a visible and near-infrared (NIR) VHSR MS image wereavailable.

The United Nations High Commissioner for Refugees (UNHCR) Bangkok o� ceprovided the author with a 1510 000 scale map of Site 2 settlement scheme dated22 January 1990. This map was produced by the United Nations Border ReliefOperation (UNBRO) Engineering Unit using Computer Assisted Design (CAD)software in combination with advanced � eld survey equipment (R. Burrows 1999,Personal communication) . The map showed how Site 2 was organized into six camps,Ampil, Banthat (for administrative purposes only), Dang Rek, Nong Chan, Rithysenand San Ro. On the UNHCR map, detailed features such as administrative centres,water sites, hospitals and sanitation centres were mapped. In addition, the areacoverage of all six camps were listed. Corresponding camp populations are listed inLynch (1990).

The source of satellite data uniquely suitable for this study is Sovinformsputnik,an organisation that markets VHSR images derived from scanned photographs fromthe KVR-1000 intelligence sensor. The KVR-1000 sensor used in this study canacquire images with two light-sensitive � lm layers: a panchromatic layer and a NIRlayer, with the NIR layer providing additional useful information for vegetationassessment (Tucker 1979). The MS spatial resolution of the image is 3.3 m, which ishigher than the resolution available from other commercially available MS satelliteimages today. Table 1 lists the characteristics of the image used in this study.

3. MethodThe VHSR MS image was geographically recti� ed based on image centre and

corner coordinates as listed by the data provider. These ground control points(GCPs) were based on the satellite’s position and listed as better than one minute-degree, i.e. better than 1.85 km by the data provider. This is a relatively low level

Table 1. Parameters of the original KVR-1000 image used in this study.

Satellite Panchromatic Infrared Image Spatialheight wavelengths wavelengths dimensions resolution

Date Time (km) (nm) (nm) (km Ö km) (m)

7 February 04535GMT 219.9 570–690 670–810 15 Ö 15 3.31989

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 5: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

Remote Sensing L etters 613

of horizontal accuracy compared to the 12 m horizontal accuracy listed for theforthcoming VHSR satellite sensors (Space Imaging 1999).

The true-coloured MS characteristics of the satellite image in combination withVHSR information facilitated visual interpretation of Site 2’s environment andvegetation. A multi-layered digital space map of Site 2’s infrastructure was createdbased on visual inspection (� gure 1 and cover).

Individual camp areas were determined from polygon vector overlays representingcamp borders from which the image analysis tool, ERMapper, provided automaticarea calculations. Based on VHSR satellite sensor data as well as in situ derived areaand in situ population estimates for individual camps, a statistical linear regression

Figure 1. Multi-layered space map of Site 2. Image copyright Sovinformsputnik, 1998.

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 6: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

E. Bjorgo614

analysis was carried out on the dataset in order to assess a potential relationshipbetween area and population.

4. ResultsFrom � gure 1 it is evident that the relatively dense vegetation in the north-west

part of the image appears to have been altered due to the construction of the refugeecamp. From in situ information it is known that housing was constructed from localmaterials such as bamboo and thatch (Lynch 1990). It is likely that these materialswere collected from this previously forested area. From the infrastructure seen in theVHSR MS image, as well as information from in situ sources, it is evident that Site 2was logistically a relatively well administered settlement (Lynch 1990, Shook 1990,R. Burrows 1999, Personal communication) .

The VHSR MS characteristics of the satellite image in combination with theUNHCR map revealed eleven water sources scattered through the site (� gure 1).The road network connecting Site 2 with neighbouring villages and main roads wasalso relatively well developed, as seen east of San Ro camp in � gure 1.

Table 2 lists area and population characteristics of the individual camps. Thetotal area of Site 2 derived from the polygon vector overlays of the satellite imageis 8 189 848 m2 , while the total camp area as listed on UNHCR’s map is8 095 000 m2 —a di� erence of 94 848 m2 (1.2%) (individual camp areas di� ered more).The total populated area of Site 2 was estimated to 4 851 388 m2 based on visualanalysis of the VHSR MS image. This resulted in an average population densityestimate of 29 522 persons km Õ 2 , given the total population of 143 221 (Lynch 1990).

As may be seen from � gure 2 and table 3, the statistical relationship betweenrefugee population and camp area is linear and signi� cant within a con� dence levelof 99%. The slope of the VHSR satellite sensor-derived population versus area andthe in situ derived population versus area is 0.020 and 0.021 respectively.

5. Discussion and conclusionThe combination of VHSR and MS information was found to be advantageous

for an environmental analysis of Site 2 and its surrounding area, with the NIRinformation facilitating recognition of di� erences in vegetation. The detailed spatial

Table 2. Areas, populated areas, population and population density of Site 2’s individualcamps. The refugee population data are based on a census dated 24 August 1989(Lynch 1990). Banthat is an administrative camp with no refugee settlement. Thepopulation density was computed from the ratio of refugee population to populatedarea.

Area (m2 ) Satellite Populated PopulationArea (m2 ) from area— area (m2 ) density

from in situ in situ from Refugee (personsCamp satellite data area (m2 ) satellite population kmÕ 2 )

Ampil 1 274 727 1 365 000 Õ 90 273 948 986 24 658 25 984Banthat 781 321 655 000 126 321 N/A N/A N/ADang Rek 1 485 999 1 220 000 265 999 580 910 19 069 32 826Nong Chan 824 442 880 000 Õ 55 558 719 867 22 256 30 917Rithysen 3 213 821 3 255 000 Õ 41 179 2 189 018 66 546 30 400San Ro 609 538 720 000 Õ 110 462 412 607 10 692 25 913

Total 8 189 848 8 095 000 94 848 4 851 388 143 221 29 522

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 7: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

Remote Sensing L etters 615

Figure 2. Linear regression analysis of (a) in situ derived refugee population versus satellitesensor-derived camp area and (b) in situ derived refugee population versus in situderived camp area. The 95% con� dence level error bars for population estimatesare shown.

information added considerable value to infrastructure analysis. The overall di� er-ence of 1.2% between the satellite-derived area and UNBRO Engineering Unit � eldsurvey area suggests that VHSR MS images are suitable for area estimation. Thestatistically signi� cant relationship between population and area derived here showspotential for use in planning and administering refugee camps with populationdensities similar to that of Site 2 (29 522 persons km Õ 2 ).

A more complicated, but potentially more accurate, method of estimating popula-tion via remote sensing is through dwelling unit counts and corresponding in situ

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014

Page 8: Using very high spatial resolution multispectral satellite sensor imagery to monitor refugee camps

Remote Sensing L etters616

Table 3. Results of linear regression statistical analysis on the relationships between VHSRsatellite sensor-derived camp area and in situ derived population, and in situ derivedcamp area and in situ derived population (see also � gure 2).

VHSR satellite sensor-derivedcamp area versus in situ derived In situ derived camp area versus

Parameter camp population in situ derived camp population

Intercept Õ 1541 Õ 2660Slope 0.020 0.021P-value 0.010 0.002Con� dence level (%) 99 99

information on the number of persons occupying each dwelling unit (Watkins andMorrow-Jones 1985). However, the spatial resolution of 3.3 m of the satellite imageused here was too coarse for the detection of individual dwelling units in the densesettlement structure of Site 2.

Imagery of spatial resolution down to 1 m in panchromatic mode and 4 m in MSmode from forthcoming satellites, such as Ikonos-2, OrbView-3 and QuickBird, haspotential for more e� cient management of humanitarian relief operations, includingdwelling unit counts in densely populated refugee camps.

AcknowledgmentsThis work was sponsored by the Nansen Fellowship Fund, Kavli Foundation,

Bergesen Foundation, Hejes Foundation, Fossums Foundation, and the US Instituteof Peace Program on Virtual Diplomacy.

ReferencesAplin, P., Atkinson, P. M., and Curran, P. J., 1997, Fine spatial resolution satellite sensors

for the next decade. International Journal of Remote Sensing, 18, 3873–3881.Bouchardy, J.-Y., 1995, Development of a GIS system in UNHCR for environmental,

emergency, logistic and planning purposes. Environment Unit Report 1-95, DPOSUNHCR, Switzerland.

Elo, O., 1996, Prevention of, preparedness for, and relief of natural disasters. In EvolvingT rends in the Dual Use of Satellites, edited by P. Gasparini Alves (Geneva: UnitedNations Institute for Disarmament Research), pp. 65–75.

Gupta, V., 1995, New satellite images for sale. The opportunities and risks ahead. InternationalSecurity, 20, 94–125.

Lodhi, M. A., Echavarria, F. R., and Keithley, C., 1998, Using remote sensing data tomonitor land cover changes near Afghan refugee camps in northern Pakistan. GeocartoInternational, 13, 33–39.

Lynch, J. F., 1990, Border Khmer—a demographic study of the residents of Site 2, Site B,and Site 8. Joint Voluntary Agency Program Report No. 1/90, American Embassy,Thailand.

Shook, G., 1990, Development of environmental health criteria for locating and assessingdisaster relief and refugee camps, using a multiattribute utility rating technique. Ph.D.Thesis, Department of Environmental Health Sciences, Tulane University, NewOrleans, USA.

Space Imaging, 1999, Space Imaging Product Catalogue 1/99. Space Imaging, Colorado, USA.Tucker, C. J., 1979, Red and photographic infrared linear combinations for monitoring

vegetation. Remote Sensing of Environment, 8, 127–150.Watkins, J. F., and Morrow-Jones, H. A., 1985, Small area population estimates using aerial

photography. Photogrammetric Engineering and Remote Sensing, 51, 1933–1935.

Dow

nloa

ded

by [

Uni

vers

itat P

olitè

cnic

a de

Val

ènci

a] a

t 11:

13 2

5 O

ctob

er 2

014